Skip to content

Pull requests: sonos/tract

New pull request New
31 Open 1,939 Closed
31 Open 1,939 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

linalg/x86_64: AVX-512_FP16 native f16 hardswish kernel (stacked on #2310)
#2313 opened May 28, 2026 by czoli1976 Contributor Loading…
5 tasks done
1
drop the unmaintained atty crate; use std::io::IsTerminal
#2312 opened May 28, 2026 by kali Collaborator Loading…
linalg/x86_64 + core/nn: fused AVX-512 RmsNorm kernel
#2311 opened May 28, 2026 by czoli1976 Contributor Loading…
8 tasks done
linalg/x86_64: add AVX-512 f16 element-wise activations (stacked on #2304)
#2310 opened May 28, 2026 by czoli1976 Contributor Loading…
4 tasks done
Runtime names
#2309 opened May 28, 2026 by kali Collaborator Loading…
linalg/x86_64: add AVX-512 f16 softmax_l2 kernel
#2307 opened May 28, 2026 by czoli1976 Contributor Loading…
3 tasks done
1
linalg/x86_64: add AVX-512 erf kernel
#2306 opened May 28, 2026 by czoli1976 Contributor Loading…
3 tasks done
linalg/x86_64: add AVX-512 softmax and max reductions
#2305 opened May 28, 2026 by czoli1976 Contributor Loading…
3 tasks done
linalg/x86_64: add AVX-512 element-wise activations
#2304 opened May 28, 2026 by czoli1976 Contributor Loading…
3 tasks done
linalg/x86_64: add AVX-512 VNNI int8 GEMM kernel (avx512vnni_mmm_i32_8x8)
#2303 opened May 28, 2026 by czoli1976 Contributor Loading…
3 tasks done
onnx: auto-enable external_state to inline single-iteration Scan for streaming RNNs
#2295 opened May 27, 2026 by czoli1976 Contributor • Draft
3
onnx,core: fuse standard ONNX LSTM cell into one LstmEpilogue op
#2294 opened May 27, 2026 by czoli1976 Contributor Loading…
onnx: handle com.microsoft RotaryEmbedding contrib op
#2284 opened May 26, 2026 by czoli1976 Contributor Loading…
arm64 SME2 SMOPA int8 matmul kernel (sme_qmmm_i32_32x32)
#2279 opened May 24, 2026 by czoli1976 Contributor • Draft
3
arm64 SDOT int8 matmul kernel (FEAT_DotProd) + PackedI8K4 packing
#2278 opened May 24, 2026 by czoli1976 Contributor Loading…
6
onnx: fix LayerNorm output dtype mismatch with F16 inputs
#2276 opened May 24, 2026 by dqj1998 Loading…
1
linalg/mmm: cache-adaptive 2D-blocking for the single-thread tile walk
#2274 opened May 23, 2026 by czoli1976 Contributor • Draft
10
linalg/arm64/sme: experimental f16 SME kernels for Apple M5/A19
#2273 opened May 22, 2026 by czoli1976 Contributor • Draft
core/onnx: fused OptGru op (draft — path-C alternative to #2257)
#2260 opened May 20, 2026 by czoli1976 Contributor • Draft
2
feat: add tract_moe_ffn operator for Mixture-of-Experts FFN
#2084 opened Mar 30, 2026 by JulienBalianSonos Collaborator Loading…
Flux schnell
#2082 opened Mar 30, 2026 by kali Collaborator Loading…
Refactor nemo examples
#2006 opened Mar 5, 2026 by kali Collaborator • Draft
tolerate / in id names
#1885 opened Oct 16, 2025 by kali Collaborator • Draft
[DRAFT] Feat/vptq proto
#1627 opened Jan 29, 2025 by JulienBalianSonos Collaborator • Draft
AVX512 kit
#1567 opened Oct 29, 2024 by kali Collaborator Loading…
ProTip! Add no:assignee to see everything that’s not assigned.